Quantum Reservoir Parameter Estimation via Fisher Information

Authors : Ufuk KORKMAZ, Deniz TÜRKPENÇE

Pages : 388-396

Doi:10.16984/saufenbilder.1018716

View : 19 | Download : 9

Publication Date : 2022-04-30

Article Type : Research Paper

Abstract :In this study, we show that as a result of weak interaction of different information environments structured with a single probe qubit, these environments can perform binary classification of the information they contain. In this way, we refer to these environments as quantum information baths because they consist of sequences of identical qubits in certain pure quantum states. A micro-maser like master equation has been developed to clearly describe the system dynamics analytically and the quantum states of different information reservoirs. The model can also be treated as a quantum neuron, due to the single-qubit probe that makes a binary decision depending on the reservoir parameters in its steady state. The numerical results of the repeated interaction process based on the divisibility and additivity of the quantum dynamic maps are compared with the analytical results. Besides being a single quantum classifier, the model we present can also serve as a basic unit of a quantum neural network within the framework of the dissipative model of quantum computing.
Keywords : Binary classification, information reservoir, collision model, quantum Fisher information